Nutribullet? No, another Blender

General Discussion
1

Something approaching 20 years ago, I became aware of an attractive seeming 3D software suite called Blender. It promised many things, and it was free. I downloaded it, and I tried to learn to use it. I tried really hard.

It was a bloody nightmare. The UI was a mess. Nothing worked properly. There were thousands of possible perrmutaions of settings, each of which would change the behaviour of all the other settings. Worst of all, there was absolutely zero documentation. I was faced with this mind-blowingly complex software, dealing with obscure and esoteric concepts - with no guidance at all.

It was a bit like trying to learn Swahili by listening to people speak it on the radio. Without some kind of reference, it was just meaningless.

After a couple of months, I gave up. Blender would either die, or it would improve. I would come back and try again if things got better.

Maybe ten years ago I tried Blender again. This time there was documentation, and there were tutorials on the web. People had made movies with Blender. I was hopeful, and I gave it a really good shot.

It turned out that most of the available tutorials were for older versions of Blender, and Blender didn’t work that way any more. It’s incredibly frustrating to plough your way through hours of video tutorial, taking notes as you go, only to find that the buttons the instructor was pressing don’t exist any more.

Not only that, but the documentation wasn’t written by the developers. It was cobbled together piece-meal by enthusiastic amateurs - some of whom understood how the software worked, and many of whom who didn’t. A lot of the documentation was out of date. Some of it was just plain wrong, and had never been right, and none of it was written by people who had even a modicum of communication skills.

It was a nightmare again. This time, it was like trying to learn Swahili, but with an instruction manual - except the instruction manual was written in Japanese. Probably by Eskimos.

I gave up again. Perhaps Blender would improve. Maybe I would try again when it did.

Well, we now have Blender 5.1, and it’s fully documented, by the development team. The UI has undergone significant changes, and (I am told) Blender has improved dramatically.

So here I am again, having another go at learning Blender.

Wish me luck.

10 Likes
2

One very important thing when using blender is to have a focus on the desired result. You don’t “learn blender”. The thing has such a ludicrous amount of options that that is a fools errant.

But you can learn to do specific things in blender while remaining blissfully ignorant of the other 2000 options to do the same thing. I find that works reasonably well.

10 Likes
3

Well the goal (for now) is to build an island. If you’ve played Myst or Dear Esther you get the idea.

I want it to be as realistic as possible, by modern standards. I want it to be fully walkable, and maybe 5 square kilometres in size. I want it to have varied terrain - beaches, cliffs, grassland, lakes, pools, streams, seashore, forest, hills and mountains. I want to populate it with assets - trees, shrubs, plants, flowers, buildings (ruins?) rocks, walls, fences etc. that I have made myself.

For the moment, I am content for the island to be abandoned. I don’t, as yet, have any desire for NPCs or character animation - although wildlife might be nice.

Ultimately, it would be nice to export the whole thing to an engine like Unity or Godot. I’m aware that would cause problems with shaders.

I don’t see it as a short-term project. I’m happy to spend years over it, off and on, as long as I’m making progress.

9 Likes
4

Good luck.

8 Likes
5

Funny you post this. I recently started learning blender(again) also :grinning_face: Im about 3 weeks in, spending 2-4 days a week either learning something new and/or practicing with it. As a bonus, my wife is joining me on this endeavor as she wants to learn too so that maybe we can push our side business into a different direction of modelling products in 3D for flashy eye-catching animated advertisements.

For some background, I can 3D model but with softwares other than blender as I mostly needed a software that was accurate for precision parts I 3D print for around my house for repairs or custom needs. I also sell 3D printed goods on the side in which I have made a couple hundred sales in the span of about a year.

Blender was a huge hurdle 10 years ago but it was doable from what friends would tell me. However, I just didnt see a reason to jump through hoops to make what sketchup could do quicker(for precision parts). Sketchup WAS free till they were bought out and they slapped a hefty subscription model to it. There is a free web version, fortunately, but they stripped it of all the most useful features and removed the use of any plugins. Still, I was getting by with that for a few years, even sold a few thousand dollars worth of 3D printed things with it, but now I have a reason to move on to Blender as there are shapes and angles I cant do in Sketchup without the plugins. And the free web Sketchup doesnt allow large complex models before it crashes.

So, here I am, learning Blender again after about 10 years being away from it.

What ive learned in my years of 3D modelling is not to think “I want to learn the software”, but instead think “I want to learn only what I need to make what I want with the software” because you’ll never realistically learn all of Blender, its too much for one person and many things you could learn may never be used again. Even professionals dont know how to fully use Blender, only what they need for their work. And that may be only a handful of things.

So if you want a good place to start:

The Blender Guru guy on youtube. I never took him seriously in his older Blender videos but he has an updated Blender 5.0 series now(youtube “Blender Donut Tutorial Part 1 2026”). He has you make a donut and a mug and its fully up to date. He’ll now pitch his paid Blender lessons in his videos but I dont know the quality of those as its new and I havent tried it so just ignore it unless you want to take a chance on the guy.

So, a donut and a cup is not an island… BUT most of the tools he will use to make these things will apply to the majority of anything you intend to make, including an island, specially the parts where he gets into making lumps and mounds on the donut to make imperfect surfaces like an island would have! So I strongly suggest starting with the Blender Guru for the fundamentals, and after each video, create something, anything, using what you just learned to help drive the concepts in your brain and muscle memory. For me, it was a crude looking fish, a bottle, a pastry, a simple spaceship, etc.

After those videos, you could make an island probably just fine, but I would first watch how other make them first just so that you can get an idea of what would be the best angle of approach. I believe there is an addon for making islands that can save you a lot of time, then using what you learned with the donut tutorial, you can then manipulate the island for the precise shape you are looking for.

After my wife and I are done grasping as many basic concepts as we feel comfortable with to make common products(like a shampoo bottle, funky glass shapes, lamps, etc), we intend to move on to a cheap Unity course that focuses on setting up scenes, simple animations, lighting, texturing, rendering, etc. but that may be a few weeks from now.

8 Likes
6

Sounds like a fun project, keep us posted! Good advice to focus on specific goals, not the whole app.

I had exactly the same first reaction to Blender… I had no project to go with it (I was part of a game dev group but I was not the modeller), so I had no impetus to use it for more than converting models. I drew a different conclusion than you: a) I gave up on wanting naturalistic models, and b) I contributed heavily to wikis that explain software/development/games to people!! Hopefully I helped…? :face_with_hand_over_mouth:

I assume the original GUI just grew in complexity, and the dev knew where everything was, and didn’t realise what it looked like to newcomers? And probably he didn’t study “UX design”. And later contributors helped him improve the user experience?

7 Likes
7

I tried your advice with the Blender Guru guy.

It’s not exactly the island I was thinking of…

island2
island21844Ă—895 435 KB

8 Likes
8

As I understand it, the developers of Blender were originally a commercial animation studio, and Blender was their own in-house, self-developed software. As such, I don’t think they particularly cared what it looked like, or whether outsiders could understand it. They understood it, because they made it. It wasn’t developed for anyone else to use, and there was no need to document its functions.

7 Likes
9

I would like to reserve a beach chair near a green sprinkle. Is there a charge if I nibble on the island?

6 Likes
10

That’s where things get a bit more complicated… You’ll have to make several LODs and cut the whole thing up into pieces. It’s quite a lofty goal for a learning project, I wish you luck!

6 Likes
11

As I said, I’m not in any hurry. If it takes years, that’s OK, as long as I’m making progress.

(edit) I’ve been trying to use this thing for 20 years. A few more won’t hurt.

6 Likes
12

A little demo I found - assets (trees, plants, textures etc) created in Blender - the whole thing run in Godot Engine. If I can create something that looks this good, I’ll be happy.

6 Likes
13

This admittedly gave me a good laugh :joy:

Might need to finish the 2026 tutorials though to see what im talking about.

Im still new at Blender myself, but in my head what I was thinking is to draw the map of your island on paper, then cut it up into smaller, more manageable pieces. (The goal would be to make the island in pieces and stick them together afterwards)

Then, in blender, take a cube and stretch it into a wide plane to cover the piece of the island you are working on. From there, use what you learned for the icing of the donut to make the floor uneven. Then find and add a dirt texture to it like you would from the later parts of the tutorial (the guy pushes his own company for textures but he gives some free textures for learning purposes, you can get them from anywhere though) as this should then give your uneven ground a realistic dirt look.

6 Likes
14

Making the island is a long-term goal. Even if I can create it in Blender, it seems unlikely the entire design would transfer directly to a game engine. At the moment, I’m content to create small scenery assets, and become familiar with modelling, textures and shaders. I’ve made a few of my own textures - I’m reasonably familiar with 2D graphics - but making shaders is new to me, so that’s proving to be be an adventure.

The way I see building an island is by prototyping the shape in Blender, then re-creating it in a suitable game engine. I don’t imagine it would be possible to transfer the whole thing over. Assets such as buildings, vegetation, machinery etc can be stored as models, and transferred. Textures will transfer directy, but shaders will need some work.

As I’ve said, it’s not a short-term project.

5 Likes
15

What do you mean by that?

What Jupiter is referring to is a best practice. Don’t make large things as one large 3D object. Similarly, also a large core object with small addons is suboptimal. The 3D algorithms are optimised to hide (i.e. not calculate) terrain pieces that are out of sight. And that’s only possible if the models are pieces. If there is one large piece that is always partially visible, it must be calculated all the time, which slows everything down.

(Am I right in assuming that’s what you meant, Jupiter…?)

5 Likes
16

I mean that if, in Blender, I buid a 5 kilometer by 5 kilometer island, with all its attendant hills, cliffs, rocks, mountains, streams, beaches, trees, shrubs, plants, pathways and buildings, it seems unlikely that I would be able to copy the whole thing as a single .blend file into say, Unity, and expect it to work.

I expect, rather, to have to create individual trees, individual buildings, individual rocks etc, and copy them over as single models. The landsscape I would expect to have to create in Unity itself.

3 Likes
17

I see! That is true as well. Make each tree, each rock, etc. a separate model. That allows you to reuse them flexibly as well. I don’t know what tools Unity has for large landscapes.

Does your island have caves or overhangs? I don’t know how much you already know.

There are different types of terrain

  • Voxel (like NMS, likely overkill for you)
  • Heightmap
  • chunked Blender meshes

Shooters and RTS often use heightmaps as floors, and put trees, buildings, and bridges on top. They just need plains, hills, and maybe a riverbed somewhere. The advantage is that walking on that kind of ground is fast to calculate for a game engine. The disadvantage is, that it cannot contain overhangs, tunnels, or caves (but unless you’re making a game about caves, that’s usually no issue.) This also explains why in many games, there are weird transitions between outside and tunnels/caves. :slight_smile:

And meshes can have any shape and need no floor (can be hollow asteroids in space, floating spaceship debris, islands full of tunnels), but for large maps, you need to do more own work to define how things are optimised, broken up and unloaded when they are out of sight of the player.

No need to calculate blades of gras, wallpapers, and flowerpots when they are 20m away. No need to calculate that room when you’re in this room, that floor when you’re on this floor, those houses when you’re in this street. But if you can see parts from afar (think of house facades in cities, or doors in hallways, or hills on islands), then you need to load those parts separately (minus their flowerpots), that needs a bit of planing.

5 Likes
18

Yea. I probably should have added the optimisation part too because, unless you have a beefy PC, blender might get bogged down if you do it in one go too.

But I was mainly focusing on the part where its easier and less intimidating to start with a small portion of an island first. Then work on the next “piece” that connects to the first piece. Also helps if you have a weird geometry issue, it’ll be isolated to a single asset rather than the whole thing.

I cant speak much about Unity but I played around with Unreal engine a bit more than Unity(nothing serious, just for fun) and I didnt have any issues importing random things directly into Unreal from Blender. That was some time ago and the objects were just the simple default shapes stretched out to different sizes with a few random assets I found on the internet but I remember following a tutorial and it was pretty simple. However, I cant speak on shaders within Unreal. I never got that far. I just made a dumb obstacle course for the fun of it one day. But I didnt have any issues with any of the assets. Was pretty strait forward.

My understanding is that the recommended approach to making a game is to just make your assets in Blender with either Unity or Unreal in mind from the start. I know that might be obvious, but I guess its the best answer I always would get when I would ask about this.

4 Likes
19

After reading all the comments ive come to the conclusion this isnt the place to push my nutribullet stock or sign you up to my blender based nutrient MLM pyramid scheme.

If i sign up 1000 more people i can finally be my own boss! :stuck_out_tongue:

Been dabbling in blender again to brute force my nms base builder idea. Dabbled with it again during lockdown when also using photogrammetry software to insert classic ps1 characters into highly detailed models of some nearby landmarks.

Look forward to seeing your journey and the island that documents your progress

4 Likes
20

Currently deep-diving into textures, shaders, and materials. Things have moved on a lot since my days playing with Source Engine.

Fortunately, most of it can now be done via nodes. Not so much being utterly mystified by raw C++ code.

4 Likes